Think big — think omics
- 23 Downloads
Traditionally the laboratory diagnosis of inborn errors of metabolism largely relies on targeted hypothesis-driven measurements of metabolites in body fluids. The hypothesis is usually based on the phenotypic characteristics of the patient, defined through deep phenotyping or phenomics. The biochemical phenotype, or rather the ‘metabolite profile’ reflects both endogenous factors such as genotype(s), gene expression, the different chemical reactions taking place in the body, as well as exogenous factors such as dietary habits, drug metabolism and the microbiome. The last decade has seen the emergence of untargeted, hypothesis-free measurements. NMR spectroscopy and mass spectrometry in combination with different separation techniques (LC, GC or CE) pave the way to an important next step in our understanding of inborn errors of metabolism. The aim of untargeted metabolomics is to identify and measure as many metabolites as possible, including unknowns, to generate the metabolic fingerprint characteristic for a biological sample and indicative for genetic defects influencing human metabolism. Typically, untargeted metabolomics techniques show more than 10,000 “features” in a single body fluid sample. These are big data! Bioinformatic- and chemometric techniques are required to reveal the relevant diagnostic features, which in turn help identify the specific aetiologic condition in the individual patient. This approach is rather different from classical metabolomics studies that usually describe the comparison between a patient group and a control group. This issue of the journal illustrates that “next generation metabolic screening” (NGMS) techniques enable us to use untargeted metabolomics for diagnostics in a clinical setting (Coene et al 2018). Untargeted metabolomics techniques are approaching maturity. It is demonstrated in this special JIMD issue that untargeted metabolomics techniques can identify the vast majority of IEM-diagnoses as reliably as the conventional techniques. In addition, many examples from NMR spectroscopy and untargeted MS analyses have already emerged that unravel the identity of hitherto unknown inborn errors of metabolism (van Karnebeek et al 2016; Pol et al 2018). These techniques also have the power to discover novel biomarkers even in well-known and in-depth studied inborn errors of metabolism (Abela et al 2016; Abela et al 2017).
Although there is progress, it is at the same time clear that we are far from understanding the full complexity of the many metabolites that float in our body fluids. The data comprise many “features of unknown significance or identity” (FUS). Analogous to deciphering the significance of genetic variants of unknown significance, national- and international collaboration between metabolic groups will be pivotal to bring our understanding of human metabolism to the next level. Databases like the Human Metabolome Data Base (HMDB) curated by Professor David Wishart at The Metabolomics Innovation Centre (TMIC) in Edmonton Canada are a perfect vehicle to store our new metabolic knowledge and open it up to the IEM- and broader scientific community (Wishart et al 2018). Input from groups that work on the microbiome and groups that study how food components are metabolised will be invaluable. Central IEM knowledge repositories like IEMbase map to both HMDB and human phenotype onthology (HPO) and offer a diagnostic aid for many genetic metabolic centres and clinical communities seeking support in the diagnosis of IEMs (Lee et al 2018).
However, there is a new frontier still. For decades we have focused on the metabolites that escape the cell. The intracellular metabolome was left largely untouched for diagnostic purposes, not to mention the metabolome of the many cellular organelles. Samples, such as fibroblasts and blood cells, surely can shed new light on the forgotten intracellular metabolome. The metabolome of these cells will be a first step towards analysis of metabolites involved in tissue specific or cell type specific metabolic pathways. The majority of organ specific reactions occur in the liver, kidney, brain and colon.
Unfortunately, no single analytical technique will be able to reveal the complete spectrum of metabolites in body fluids or cell extracts. That is why the analysis of lipids is set apart from the untargeted metabolite analysis. For many decades doctors have been measuring cholesterol and its sub-fractions in combination with triglycerides primarily for vascular risk analysis. Inborn errors like Barth syndrome, MEGDEL syndrome and many others have made it clear that knowledge on a vast array of complex lipids is crucial for our understanding of inborn errors of lipid metabolism (Wortmann et al 2012; Garcia-Cazorla et al 2015). Some experts suggest that the human body may contain >5000 different lipid species. This has given rise to the development of untargeted lipidomics methodologies. They hold a big promise for the future of the metabolic field.
The last three decades have brought us the many forms of congenital disorders of glycosylation as a new kid on the metabolic block. Initially using transferrin isoelectric focusing of blood samples as a diagnostic fishing tool, untargeted glycomics techniques are now slowly emerging. These will bring techniques for the analysis of glycoproteins and glycolipids. Also, in these “glyco-IEMs” mass spectrometry techniques will allow us to make a huge step forwards. MS-analysis of intact transferrin is already feasible and brings more detailed information regarding the glycan composition than isoelectric focusing. Further technical innovations quantify individual glycan species after their release from proteins. However, the best is yet to come. In the near future it will be possible to use MS for the analysis of all glycopeptides deriving from all glycoproteins after tryptic cleavage of the proteins in a sample. This will unravel which glycan is at each specific glycosylation site of a protein. Of course, there will be no need to focus exclusively on the plasma glycome because the intracellular glycome will be as relevant for specific IEMs.
Metabolomics, glycomics and lipidomics currently form the corner stones of untargeted techniques instrumental for the IEM-field. Of course, we should not forget the post-synthetic protein modifications for which proteomics analyses may become important. Also, the proteome may again come into play when analysing protein-protein interactions, the interactome or composition of protein complexes. Complexome profiling, or complexomics, is around but is not used heavily yet (Guerrero-Castillo et al. 2017).
This JIMD issue on –omics paints a picture of a new dawn for clinicians and laboratory specialists in the metabolic field. Untargeted “new school techniques” are emerging and have the potential to replace several of the traditional targeted diagnostic techniques. However, "Don't throw away your old shoes before you have new ones" as we say in Dutch. Validation of untargeted techniques has several aspects. Apart from the analytical chemistry validation, a robust and well validated chemometric pipeline should be in place to interpret the big data to be sure not to run into artefacts or false discoveries and to be capable of squeezing out the correct and relevant diagnostic features. Consistent monitoring of QC samples within run and between runs must guarantee the quality of the analytical performance of the system. Yet there are many inborn errors of metabolism and based on a clinical validation the metabolic laboratory specialists must be sure which diagnoses an untargeted platform can make and which diagnoses will be missed. Analysing ERNDIM samples on these platforms will significantly contribute to the overall quality control of these procedures.
Needless to say, the diagnostic landscape is changing at a rapid pace. The previous 5 years have seen whole exome sequencing (WES) mature. It is now central on the clinical stage thus giving a new impetus to the field of genetic and metabolic diseases. Many novel diseases, sometimes with signs and symptoms that have never been seen before in any IEM, are uncovered by using exome sequencing. The genetic basis of new phenotypes is unravelled at a high pace. In some cases, novel phenotypes connect to defects in well-known genes thus expanding our knowledge of the phenome. Whole genome sequencing will further accelerate this process. Have we found the Holy Grail? Possibly, but for sure our understanding of how things actually work in our cells at the level of enzymes, transporters and metabolites will certainly remain essential to be able to develop efficient therapy strategies and to follow-up the patients in the natural course of the disease and after therapeutic interventions. Detailed knowledge on the pathophysiology will enable more accurate prognostication as well as the development of novel therapeutic strategies. Ideally the metabolic laboratory should develop next door to or in the genetics laboratory to ensure optimal cross talk between the laboratory specialists evaluating these data. The metabolic laboratory actually is developing as a functional genomics laboratory. The metabolic laboratory may help interpret variants of unknown significance (VUS) found in WES-analysis when these occur in genes for transporters or enzymes. The diagnostic future may be that WES/WGS and untargeted omics techniques are applied simultaneously in the diagnostics of an individual patient. Interpreting these big data in concert in an integrated fashion with input from clinical and laboratory specialists as well as basic scientists is essential.
The current era is an interesting time in genetic metabolic diseases. The future looks bright. There is considerable research to be conducted, and many new techniques and discoveries are coming on the stage or within reach. This issue of JIMD paints a picture of this future and gives you the state of the art for phenomics, metabolomics, glycomics, lipidomics, transcriptomics and fluxomics as well as its integration with genomics and the role of the clinician in this big data era.
The editors of this special issue hope that the manuscripts in this issue will inspire you to enter this new world of –omics, and that each of you will start initiatives and collaborations to bring the field to a next level enabling the comprehensive molecular and physiological understanding of whole body metabolism. Patients with inborn errors of metabolism will surely benefit from such efforts. In future issues of the JIMD we welcome your comments and feedback to this special issue and we welcome your omics manuscripts and contributions.
- Coene KLM, Kluijtmans LAJ, van der Heeft E et al (2018) Next-generation metabolic screening: targeted and untargeted metabolomics for the diagnosis of inborn errors of metabolism in individual patients. J Inherit Metab Dis. https://doi.org/10.1007/s10545-017-0131-6
- Guerrero-Castillo S, Baertling F, Kownatzki D et al (2017) The assembly pathway of mitochondrial respiratory chain complex I. Cell Metab 25(1):128–139. https://doi.org/10.1016/j.cmet.2016.09.002
- Lee JJY, Wasserman WW, Hoffmann GF et al (2018) Knowledge base and mini-expert platform for the diagnosis of inborn errors of metabolism. Genet Med 20:151–158. https://doi.org/10.1038/gim.2017.108